How to avoid creating products that no one wants with JTBDs

In part 1 of this series, we focused on what is a 'need' and how to get a glimse into what people want by using situational segments.

To get the real picture why people buy what they buy, however, we need to talk more about jobs-to-be-done (JTBD).

If you're familiar with the jobs-to-be-done framework, you probably know there are multiple "sub-frameworks" on how to identify what your users are really trying to accomplish.

I've taken all of these "sub-frameworks" combined the best parts to come up with a more scientific way help to identify why people buy your product (SaaS, software, etc.)

Here are some jobs-to-be-done examples

Get a job. Slow my skin from aging. Find qualified workers. Learn a subject. Discipline a child. Manage financial security.

What do all of these have in common?

They’re all jobs-to-be-done that people are trying to accomplish.

Your product is just the current thing they “hired” to get that done.

Here’s a good template to express jobs-to-be-done (notice the examples above follow that template):

The “action verb” part starts some action that people are trying to accomplish. Are they trying to discover, achieve, learn, remember, confirm, maintain, experience, coordinate, remove, find, share, make, correct, fix, obtain, create, get, plan, stop, demonstrate, help, teach, prepare, identify, detect, prevent, understand, improve, determine, provide, update, protect, develop, slow or verify something?

Notice these are all action verbs (you can see a bigger list of action verbs here).

The second part is about the "objection of action": What are you trying to manage? Personal finances. That's the objection of action.

The contextual clarifier describes the circumstances and conditions under which the job needs to be done.

If you've read my post on trigger events, you'll already be aware of some context clarifiers. People need to manage their personal finances "at home", for example. "At home" is the context clarifier.

Is the job-to-be-done worth solving?

So many founders solve real problems that people aren't really looking to solve.

After you're aware of the jobs-to-be-done your product solves, we need to dig deeper and see if those jobs are worth solving in the first place.

I've used some questions from a jobs-to-be-done framework called outcome-driven innovation to try to get a clearer picture.

Let’s say you’re a reputation management software and the JTBD you have is “detect when a competitor said something bad about my company online”:

• "Detect" = action verb
• "a competitor said something bad about my company" = action verb
• "Online" = context clarifier

Take 5 people and ask them the following 2 questions (ask them to rate their answers from 1 (not important at all) to 5 (extremely important):

For detecting when a competitor said something bad about your company online:

• How important is this to you? (1 = very unimportant; 5 = extremely important)
• How satisfied are you with the ability to get this done? (1 = very unsatisfied; > 5 = extremely satisfied)

You could also specify what 2, 3 and 4 mean:

• 2 is "somewhat important"
• 3 is "important"
• 4 is "very important"

With satisfaction just replace "important" with "satisfied".

Don't overlook the satisfaction part. Most founders will focus on the importance on something and overlook satisfaction.

Some goals are important, but can be easily satisfied by a simple Google search. Like finding out the current weather.

Once you ask a few people and get several ratings, you calculate the opportunity score for that goal, using this formula:

opportunity score = importance + (importance - satisfaction)

Suppose people rated something “5” for importance for “2” for satisfaction, the opportunity score would be 5 + (5 - 2) = 8.

What if the reverse was the case (2 = importance, 5 = satisfaction)? The score in parentheses cannot be a negative number, so in case that happens, use 0.

So with this example will get: 2 + (2 - 5), which is not a negative number but 0.

The opportunity score formula will tell you **which job-to-be-done is worth focusing on (the higher the score is, the better). **

If you did this and got low opportunity scores, it might be a good idea and get to square 1 and try to find new jobs-to-be-done your product could satisfy. How? Interviews.

How to discover new jobs-to-be-done with interviews

Before we start: Interviewing is just one way to find jobs-to-be-done. You can also observe people, use focus groups or just participate in the environment they’re in.

The method you use for gathering data is not as important as the type of information you’re looking for.

You need to stay away from focusing/asking questions related to a product/solution, instead focus on the job people are hoping to get done. The goal is to learn about peoples’ purchasing process, as well as any internal/external influences, not whether your product/service has fulfills their need or how they feel about a particular provider.

If you've read "The Mom Test", you'll know what I mean.

Here are some questions you can ask (I’ll use a tax management software as an example):

• What are you trying to achieve when managing your taxes? What goals does our service help you accomplish?

• What do you hope to accomplish by using our software?

• What problems are you trying to resolve or prevent?

• What is it that you are trying to determine or decide?

• What would the ideal product/service help you to accomplish?

• When was the last time you used [your/competitor product]? Get them to talk about the circumstances around the purchase.

You may notice that some of these questions start with the product (like what would the ideal product help you accomplish). Using this framework, it’s okay to use a product-specific question but just as a starting point to get to the jobs-to-be-done.

At the end of the day, it’s all about what you focus on. Compare: “What would the ideal product help you accomplish” vs. “What features would you like in your ideal product.”

After you finish with the interviews, formulate peoples' goals using the jobs-to-be-done template above.

Be sure to ask for validation, if a customer tells you a story that points that they’re trying to, say, determine the lowest price of a product, then ask explicitly, “so you’re trying to determine the lowest price for a product, right?” to validate the goal.

Here are some useful tips when doing interviews:

1. The focus of the whole interview should be around what the customer is trying to accomplish, not what they’re currently doing with a product.

A valid job-to-be-done step that focuses on what the customer is trying to accomplish is “place an order”.

An invalid step that outlines what they’re currently doing is “call a supplier to place an order”.

Your JTBD shouldn’t hint in ANY way on HOW something is being accomplished, only on WHAT that person is trying to get done (not to say this is totally useless, but we're trying to follow a specific framework here to avoid false data).

Again, you might ask people questions related to what they’re doing currently to solve a problem, but use that as a starting point to find out what they’re trying to accomplish (the problem) by the solution. .

2. Focus on recent behavior, not on anticipated behavior.

JTBD thinking is against asking people to predict their opinion/behavior. For example, instead of asking people what competitor product do they plan to buy, ask what product did they buy in the past 3 weeks and go from there with the questions above.

3. There is always some sequence of actions that lead to the event of purchasing a product.

You need to “rewind time” and understand those actions, and the questions above will help you do that.

Also try asking what ELSE are people trying to/would like to accomplish before/during/after using your service. If a person tells you about a certain situation, ask “As a result of what did this occur?” to get one level up.

4. To discover several JTBDs from a single customer, ask: "If you got the product/service at a different time, how would the circumstances differ from now?"

Expect to get a different story when asking this. This is the advantage of using situational vs. demographic targeting; with situational targeting the interview can change depending on the situation; with demographics targeting there aren't much variables to change.

5. If people start talking about specific services and what they’re currently doing, redirect their attention to focus on their goals and what they’re hoping to accomplish.

Problems, not solutions.

6. Try to contradict what people say with their recent behavior. If they say that quality is important to them, ask if any of the products they recently bought in your category are less than top quality.

Look for inconsistencies of what people say and do. If they planned to buy product X but ended up with Y, probe around that.

7. Get people to describe things as concretely as possible.

If a customer says that “it’s a waste of time waiting in line”, ask after how many minutes exactly it began to feel like they were wasting their time.

Or if she says that she wants a more smooth experience, ask how would a smooth experience look like.

Always get the concrete details. Different people have different definitions of words like "easy to use", "reliable", "comfortable", "high-quality" or "durable" (one survey found that people have 21 different ways in which they define “easy to use”).

Always, always ask people what do they mean by reliable/comfortable etc. when they mention some of these words. Press them to explain more. Ask “What makes it easy to use?”, “Easy to use compared to what?”, “What else did you try that was easy to use?”

How long should interviews take? Typically anywhere from 15 minutes to an hour. It really depends from case to case, the complexity of the product etc.

How many people to interview at once? One? Many as a focus group? The method for gather information is not as important as knowing the type of information you’re looking for from customers (looking for jobs they’re hoping to get done vs. feature requests)

Be patient with yourself because interviewing is like any other skill. You didn’t give up the first time you encountered a software bug, did you? You’ll become better the more you do this.

Are you talking to the wrong people?

According to job-to-be-done strategists, It’s important you talk to your average user, not the lead ones that use your product most frequently.

Also, make sure you are also talking to the job executor, the person who will use your product.

In B2B, there are often two different people responsible for buying and using the product.

The CTO might be the one responsible for purchasing technology courses, but the tech support team is the one watching (using) them.

Talk to both of them, because they use different metrics to evaluate your product and have different jobs-to-be-done.

Without the decision maker, you cannot sell your product to the job executor.

The main point is to not let the purchaser talk in the name of the end user if you want to improve your offering.

You don’t want to talk to the husband that went to the grocery store if all he did was take an order from his wife to get a wishlist of her items.

Give priority to these 4 categories of customers:

• People who recently bought your/your competitor product

• People who recently quit using your/your competitor products. Asking a question like “What did you really think you were buying?” is a good idea here.

• People who recently changed to a competitor or who came from a competitor to you.

• People who use your and your competitor product (different products in the same category). They will provide useful insights on how they’re using similar products for different purposes.

You can also maximize the information you get from a single customer by asking them to tell you the stories around different times when they purchased a product in the same category (the stories will usually be quite different).